agentsclimarketplace

Agentverse image gen

Skill fetchai/agentverse-skills/skills/agentverse-image-gen

Agent skills for interacting with Fetch.ai's Agentverse — portable SKILL.md format for Claude Code, Codex, Copilot, Cursor, Gemini CLI

Install
npx -y skills add fetchai/agentverse-skills --skill agentverse-image-gen

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Generate images using AI agents on Fetch.ai's Agentverse. Sends a text prompt to an image generation agent and returns the generated image URL. Handles agent discovery, relay deployment, and response parsing automatically. Requires AGENTVERSE_API_KEY env var. Use when asked to generate, create, or make an image via Agentverse.

The file declares its own license as Apache-2.0. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

4.6 KB, as published. Nobody here has run it

Agentverse Image Generation

Overview

Generate images by sending text prompts to AI image generation agents on Agentverse. Returns a public URL to the generated image. Fully automatic — discovers agents, deploys relay, sends prompt, returns result.

When to Use

  • User asks to "generate an image using Agentverse"
  • User asks to "use an AI agent to create a picture"
  • User wants AI-generated art via the Fetch.ai agent network
  • User says "make me an image of..." in context of Agentverse

Prerequisites

  • AGENTVERSE_API_KEY environment variable set
  • Python 3.8+ with requests

Quick Steps

1. Generate an image

python3 scripts/generate_image.py \
  --prompt "A futuristic AI robot launching a rocket from a launchpad, cyberpunk style, neon colors" \
  --wait 60

2. With a specific agent

python3 scripts/generate_image.py \
  --prompt "A dragon made of circuit boards on a Tokyo rooftop at sunset" \
  --agent "agent1qdynamic8lgnax37n20296xr4kcfllahlnse7gy5mrkdt4q9v9h06qkmclkl" \
  --wait 60

3. Parse the result

When the image comes back as a CDN URL (e.g. Cloudinary), image_url is already browser-openable:

{
  "status": "success",
  "prompt": "A dragon made of circuit boards...",
  "image_url": "https://res.cloudinary.com/fetch-ai/image/upload/v1776700090/dalle3-agent/0932ec14-...",
  "metadata": {},
  "target_agent": "agent1q0utywlfr3dfrfkwk4fjmtdrfew0zh692untdlr877d6ay8ykwpewydmxtl",
  "relay_agent": "agent1q...",
  "wait_time_seconds": 35,
  "all_responses": [{"type": "text", "text": "Generating..."}, "..."]
}

When the image comes back as an agent-storage:// URI, the response also includes public_url — a direct HTTPS URL you can open in a browser or download:

{
  "status": "success",
  "prompt": "A sunset over Tokyo",
  "image_url": "agent-storage://https://agentverse.ai/v1/storage/47cecfda-8d6f-4c21-80ab-10c4d8d052bb",
  "public_url": "https://agentverse.ai/v1/storage/47cecfda-8d6f-4c21-80ab-10c4d8d052bb",
  "metadata": {"mime_type": "image/png", "role": "generated-image"},
  "target_agent": "agent1q0utywlfr3dfrfkwk4fjmtdrfew0zh692untdlr877d6ay8ykwpewydmxtl",
  "relay_agent": "agent1q...",
  "wait_time_seconds": 45,
  "all_responses": [...]
}

Use public_url (when present) to open or display the image. Fall back to image_url for CDN-hosted images.

How It Works

  1. Discover agent: If no --agent specified, searches for image generation agents
  2. Deploy relay: Creates/reuses a hosted agent as a message relay
  3. Send prompt: Sends ChatMessage with your prompt as TextContent
  4. Wait for generation: Image agents typically take 20-45 seconds
  5. Extract URL: Parses ResourceContent from response logs
  6. Return: Outputs JSON with the image URL

Default Agent

If you don't specify --agent, the script uses the well-known Fetch.ai DALL-E 3 image agent:

agent1q0utywlfr3dfrfkwk4fjmtdrfew0zh692untdlr877d6ay8ykwpewydmxtl

This agent generates images using DALL-E 3 and returns them as Cloudinary URLs.

Timing

  • ACK: ~1 second (confirms message received)
  • Image generation: 20-45 seconds (model inference + upload)
  • Recommended --wait: 60 seconds (safe margin)

Edge Cases

  • Timeout: If no image after 60s, the agent may be overloaded — retry later
  • Text response instead of image: Some agents respond with text first, then image. Increase --wait
  • Agent offline: Use agentverse-search to find alternative image agents
  • Large prompts: Keep prompts under 500 characters for best results

Example Prompts That Work Well

  • "A futuristic AI robot launching a rocket, cyberpunk style"
  • "A cat astronaut floating in space with Earth in the background"
  • "A dragon made of circuit boards standing on a Tokyo rooftop at sunset"
  • "An underwater city with bioluminescent buildings and fish"

References

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.